Controlled clinical trials have demonstrated the efficacy of reducing the blood levels of low-density lipoprotein cholesterol in reducing the incidence of coronary artery disease in hypercholesterolemic middle-aged men. However, a similar reversibility of the risk of coronary artery disease has not been demonstrated for high-density lipoprotein cholesterol elevation and triglyceride reduction. Therefore, the effect of administering 400 mg of bezafibrate retard daily versus placebo (double blind) to patients with myocardial infarction preceding randomization by 6 months to 5 years, or a clinically manifest anginal syndrome documented by objective evidence of dynamic myocardial ischemia, or both, is being investigated. Three thousand subjects (aged 45 to 74 years) are being enrolled from 19 cardiac departments in Israel, with total serum cholesterol between 180 and 250 mg/dl, high-density lipoprotein cholesterol < or = 45 mg/dl and triglycerides < or = 300 mg/dl. In addition, low-density lipoprotein cholesterol concentrations are required to be < or = 180 mg/dl (< or = 160 mg/dl for patients aged < 50 years). Patients needing lipid-modifying therapy, exhibiting > or = 1 prespecified exclusion criterion or not giving informed consent, or a combination, are not randomized. The primary end points for evaluating efficacy are the incidence of fatal and nonfatal myocardial infarction, and sudden death. The hypothesized effect of bezafibrate administration under the aforementioned protocol is to reduce an estimated cumulative end point event incidence of > or = 15% by 20 to 25% over an average follow-up period of 6.25 years, through early 1998, when the last patient recruited will have completed 5 years.(ABSTRACT TRUNCATED AT 250 WORDS)